Accurate memory for an event requires one to associate, or bind together, the different components of the event, such as the people involved, the actions that they perform, the objects that are acted upon, and the context in which the event occurs. A growing body of evidence suggests that the ability to bind information in memory diminishes with increased age in adulthood. The proposed research is designed to investigate age-related declines in one particular type of binding ability, namely the ability to bind people with their actions. This type of binding ability is crucial for memory functioning both in everyday contexts and in the less common but especially important context of eyewitness testimony. Binding people with actions is a difficult task, however, as indicated by a phenomenon known as unconscious transference. In unconscious transference, an eyewitness mistakenly associates a familiar person with the actions of another person. Given the evidence for an age-related decline in binding ability, we predict that older adults should be particularly susceptible to unconscious transference. A better understanding of the conditions under which older adults are susceptible to unconscious transference is critical to knowing when older adult testimony should be accepted with confidence and when it should be regarded with caution. The specific aims of the proposed research are (1) to develop a laboratory model of unconscious transference that will allow us to test people's ability to associate actors with actions, (2) to compare the rate of unconscious transference in young and older adults, and (3) to examine the impact of familiarity, conscious recollection, distraction, frequency of exposure, response deadlines, temporal contiguity, and photograph exposure on the likelihood of unconscious transference in young and older adults. The proposed research would achieve these aims using a method that involves presenting young adults aged 18 - 30 and community-dwelling older adults aged 60-90 with a series of brief video clips depicting a number of different actors performing a number of different actions. Participants would later be tested on their memory for these video clips. The crucial test items would involve a familiar actor performing an action that had been performed by a different actor at encoding. The false recognition of these items can be used as a measure of unconscious transference, allowing one to examine the conditions under which young and older adults are susceptible to this phenomenon.